1
Modifiable factors affecting older patients' quality of life and physical function during 1
cancer treatment 2
3
Lene Kirkhus1, 2*
4
Magnus Harneshaug1, 2 5
Jūratė Šaltytė Benth1,3, 4 6
Bjørn Henning Grønberg5,6 7
Siri Rostoft2, 7 8
Sverre Bergh1,10 9
Marianne J. Hjermstad 8,9 10
Geir Selbæk2,7,10 11
Torgeir Bruun Wyller2, 7 12
Øyvind Kirkevold1,10,11 13
Tom Borza1 14
Ingvild Saltvedt12,13 15
Marit S. Jordhøy1, 2, 14 16
17 18
1The Research Centre for Age Related Functional Decline and Diseases, Innlandet Hospital 19
Trust, P.O. box 68, 2313 Ottestad, Norway 20
2Institute of Clinical Medicine, Faculty of Medicine, University of Oslo, P.O. box 4956 21
Nydalen, 0424 Oslo, Norway 22
3Health Services Research Unit, Akershus University Hospital, P.O. box 1000, 1478 23
Lørenskog, Norway 24
4Institute of Clinical Medicine, Campus Ahus, University of Oslo, P.O.Box 1171, 0318 25
Blindern, Norway 26
5Department of Clinical and Molecular Medicine, Norwegian University of Science and 27
Technology (NTNU), P.O. box 8905, 7491 Trondheim, Norway 28
6The Cancer Clinic, St. Olav Hospital, Trondheim University Hospital, P.O. box 3250 29
Sluppen, 7006 Trondheim, Norway 30
7Department of Geriatric Medicine, Oslo University Hospital, P.O box 4956 Nydalen, 0424 31
Oslo, Norway 32
8Regional Advisory Unit for Palliative Care, Dept. of Oncology, Oslo University Hospital, 33
P.O. box 4956 Nydalen, 0424 Oslo, Norway 7 34
9European Palliative Care Research Centre (PRC), Department of Oncology, Oslo University 35
Hospital, P.O box 4956 Nydalen, 0424 Oslo, Norway and Institute of Clinical Medicine, 36
University of Oslo, P.O. box 4956 Nydalen, 0424 Oslo, Norway 37
10Norwegian Advisory Unit on Ageing and Health, Vestfold Hospital Trust, P.O. box 2136, 38
3103 Tønsberg, Norway 39
11Faculty of Health, Care and Nursing, NTNU Gjøvik, Teknologivegen 22, 2815 Gjøvik 40
12Dep of Neuromedicine and Movement Science, NTNU, Norwegian University of Science 41
and Technology, P.O. box 8905, 7491 Trondheim, Norway 42
13Department of Geriatric Medicine, St Olav Hospital, Trondheim University Hospital, P.O.
43
box 3250 Sluppen, 7006 Trondheim, Norway 44
2
14The Cancer Unit, Innlandet Hospital Trust, Hamar Hospital, Skolegata 32, 2326 Hamar, 1
Norway 2
3
Corresponding author: Lene Kirkhus, lene.kirkhus@gmail.com 4
Keywords: physical function, quality of life, geriatric assessment, geriatric oncology 5
6
Running head: Older patients with cancer, quality of life, and physical function 7
•List of where and when the study has been presented in part elsewhere, if applicable.
8
Parts of the present study were presented as a poster at the European Association of Palliative 9
Care conference in May 2019. Otherwise, this study has not been presented anywhere else, 10
but other results emerging from the same prospective, observational study has been presented 11
as referred in the present paper, and as follows:
12 13
• Kirkhus L, Saltyte Benth J, Rostoft S, et al: Geriatric assessment is superior to 14
oncologists' clinical judgement in identifying frailty. Br J Cancer 117:470-477, 2017 15 • Kirkhus L, Saltyte Benth J, Gronberg BH, et al: Frailty identified by geriatric
16
assessment is associated with poor functioning, high symptom burden and increased 17
risk of physical decline in older cancer patients: Prospective observational study.
18
Palliat Med:269216319825972, 2019 19
• Harneshaug M, Kirkhus L, Benth JŠ, Grønberg BH, Bergh S, Whist JE, Rostoft S, 20
Jordhøy MS: Screening for frailty among older patients with cancer using blood 21
biomarkers of inflammation. J Geriatr Oncol. 2019 Mar;10(2):272-278. doi:
22
10.1016/j.jgo.2018.07.003. Epub 2018 Jul 23.
23 24
3
Abstract
1
Background: Maintaining physical function and quality of life (QoL) are prioritized 2
outcomes among older adults. We aimed to identify potentially modifiable factors affecting 3
older patients’ physical function and QoL during cancer treatment.
4
Methods: Prospective, multicenter study of 307 patients with cancer >70 years, referred for 5
systemic treatment. Pre-treatment, a modified geriatric assessment (mGA) was performed, 6
including registration of comorbidities, medications, nutritional status, cognitive function, 7
depressive symptoms (Geriatric Depression Scale-15 [GDS]), and mobility (Timed Up and 8
Go [TUG]). Patient-reported physical function (PF) -, global QoL-, and symptom scores were 9
assessed at baseline, two, four, and six months by the EORTC Quality of Life Core 10
Questionnaire-C30. The impact of mGA components and symptoms on patients’ PF and 11
global QoL scores during six months was investigated by linear mixed models. To identify 12
groups following distinct PF trajectories, a growth mixture model was estimated.
13
Results: 288 patients were eligible, mean age was 76.9 years, 68% received palliative 14
treatment. Higher GDS-scores and poorer TUG were independently associated with an overall 15
level of poorer PF and global QoL throughout follow-up, as were more pain, dyspnea, and 16
appetite loss, and sleep disturbance. Three groups with distinct PF trajectories were identified:
17
a poor group exhibiting a non-linear statistically (p<0.001) and clinically significant decline 18
(>10 points), an intermediate group with a statistically (p=0.003), but not clinically significant 19
linear decline, and a good group with a stable trajectory. Higher GDS-scores and poorer 20
TUG, more pre-treatment pain and dyspnea were associated with higher odds of belonging to 21
the poor compared to the good PF group.
22
Conclusion: Depressive symptoms, reduced mobility, and more physical symptoms increased 23
the risk of decrements in older patients’ PF and global QoL scores during cancer treatment, 24
and represent potential targets for interventions aiming at improving these outcomes.
25
4 1
Introduction
2
Older adults often have complex problems, and compared to their younger counterparts, they 3
are more vulnerable, and at higher risk of experiencing a reduction in physical function, and 4
thereby functional decline and dependence, following otherwise successful treatment (1, 2). In 5
older patients receiving cancer treatment, reduced abilities to carry out daily life activities 6
reportedly occur in about 20% to 40% (3-6), and may negatively affect quality of life (QoL) 7
(7, 8). As maintaining independence and QoL are highly prioritized (9-12), decrements come 8
at high costs for the older patients, and may also significantly increase caregivers’ burden and 9
health care demands. Precise knowledge on how physical function and QoL may develop 10
during cancer treatment is therefore crucial to make treatment decisions in accordance with 11
patients’ wishes and priorities. Moreover, considering the rapidly growing number of older 12
patients with cancer and older cancer survivors (13), it is of uttermost importance to develop 13
targeted interventions that may prevent decline in physical function and QoL during cancer 14
treatment. Thus, precise knowledge on risk factors for such negative outcomes is needed.
15 16
Frailty is widely recognized as a syndrome of increased vulnerability to stressors (14). In 17
older patients with other diseases than cancer, frailty is closely related to poor QoL and an 18
established predictor of disability and dependence (14, 15). In oncology settings, a geriatric 19
assessment (GA), which includes frailty indicators such as comorbidity, polypharmacy, 20
physical, mental and nutritional deficits, is known to predict survival and side effects of 21
cancer treatment (16-19). The potential role of GA and individual frailty indicators as 22
predictors of physical function and QoL during and after treatment is scarcely investigated.
23
There are indications that impairments in activities of daily living (ADL), abnormal 24
nutritional status, and depressive symptoms may predict decline in physical function in older 25
5
patients with cancer (3, 5), but the results of the few studies available are not consistent (4, 1
20). Symptom distress may also have a substantial negative impact on physical function and 2
QoL (21-23), but for older patients with cancer, the longitudinal interrelation between 3
symptom burden, physical function and QoL during the course of treatment has not been 4
established.
5 6
We have previously demonstrated that frailty identified by a modified GA was independently 7
predictive of survival and associated with poorer physical function and more symptoms in a 8
cohort of older patients with cancer > 70 years, referred for systemic cancer treatment (24, 9
25). Addressing the same population, the aim of the present study was to identify individual, 10
modifiable factors associated with a poorer physical function and QoL during treatment. We 11
investigated the impact of pre-treatment frailty indicators on patient-reported physical 12
function and global QoL during six months after referral, and the association between these 13
outcomes and patients’ symptom reports during the same period.
14
Patients and methods
15
Patients >70 years, referred for systemic cancer treatment for a histologically confirmed solid 16
tumor (new diagnosis or first relapse after previous curative treatment) were consecutively 17
included into this prospective observational study at eight Norwegian outpatient oncology 18
clinics (two university hospitals and six local hospitals) (24). At inclusion, the patients’
19
oncologists reported cancer type according to the International Classification of Diseases-10th 20
Edition (ICD-10), stage of disease, Eastern Cooperative Oncology Group (ECOG) 21
performance status (PS), and whether patients received palliative or curative treatment. The 22
oncologists were blinded for the study specific assessments, and treatment decisions were 23
based on clinical judgment and Norwegian national guidelines. Data on administered 24
6
treatment the two first months after inclusion were retrospectively retrieved from the patients’
1
hospital medical records by checking administered infusions, prescriptions, surgical notes and 2
notes from the radiotherapy clinic. Treatment was thereafter classified as 1) curative i.e.
3
neoadjuvant or adjuvant chemotherapy, 2) palliative chemotherapy, i.e. traditional cytotoxic 4
regimens, 3) other palliative systemic cancer treatment, i.e. hormone therapy and modern 5
targeted treatment, 4) other palliative care (i.e. radiotherapy, surgery, medical symptom 6
treatment). Stage was classified as localized (I-II), locally advanced (III) or metastatic (IV), 7
and PS as 0-1 or 2-4.
8 9
Physical function, QoL and symptom assessment 10
The patients reported their physical function, global QoL and symptoms at inclusion and at 11
two, four, and six months of follow-up on the European Organisation for Research and 12
Treatment of Cancer (EORTC) Quality of Life Core Questionnaire-C30 (QLQ-C30) (26). The 13
QLQ-C30 physical function scale (PF) consists of five items: 1) any trouble doing strenuous 14
activities, like carrying a heavy shopping bag or a suitcase; 2) any trouble taking a long walk;
15
3) any trouble taking a short walk outside of the house; 4) need to stay in bed or a chair during 16
the day; 5) need of help with eating, dressing, washing yourself or using the toilet. The global 17
QoL scale consists of two items asking the patients to rate their overall health and QoL.
18
Physical symptoms are assessed on three multi-item scales (i.e. fatigue, pain, and 19
nausea/vomiting) and five single item measures (dyspnea, sleep disturbances, appetite loss, 20
constipation, and diarrhea). Fatigue was excluded from our analyses since we primarily aimed 21
at identifying factors that might be modified by targeted interventions, and since treatment of 22
fatigue generally implies identifying and treating contributing factors, including the other 23
symptoms assessed on the QLQ-C30.
24
7
All QLQ-C30 items are scored on an ordinal scale ranging from 1 (not at all) to 4 (very 1
much), except for the two items constituting the global QoL score, going from 1 (very poor) 2
to 7 (excellent). Before analyses, raw scores on all scales/items were transformed into scales 3
from 0 to 100 points (27). Higher scores on the PF and global QoL scales indicate better 4
function/QoL, whereas higher scores on the symptom scales/items denote more symptoms.
5
For all scales, a difference in scores of 5 to 10 points has been found to represent “a little”
6
difference for better or for worse for the patients, and a difference by 10 to 20 points as 7
moderate (28). Accordingly, data suggest that a 10-point change in scores represents a change 8
in supportive care needs (29). Thus, a difference of ≥ 10 points was defined as clinically 9
significant (28) 10
11
Frailty indicators 12
Frailty indicators were chosen based on a modification of the Balducci frailty criteria (24, 30) 13
and recommendations for the content of a GA (16, 18), and assessed at baseline, partly by 14
trained oncology nurses, partly by patient-report. Details of the assessment tools and 15
procedures have been described elsewhere (24) and are summarized in Table 1. Eight frailty 16
indicators were included: number of comorbidities assessed by a subscale of the Older 17
Americans’ Resources and Services Questionnaire (OARS) (31, 32), number of regular 18
medications, nutritional status using the Patient-Generated Subjective Global Assessment 19
(PG-SGA) (33), depressive symptoms using the Geriatric Depression Scale-15 (GDS-15) 20
(34), cognitive function using the Norwegian Revised Mini Mental State Examination 21
(MMSE-NR) (35), number of falls the last six months, and mobility using the Timed Up and 22
Go test (TUG) (36). The patients were asked to perform TUG at a fast pace (37). Basic ADL 23
were assessed from question 5 of the QLQ-C30 PF scale (Table 1).
24 25
8 Statistical analyses
1
The QLQ-C30 PF and global QoL scales were defined as our primary and secondary 2
endpoints, respectively. The absolute values of the patients’ scores at each assessment point 3
from baseline to six months were used in the statistical analyses. The overall course of both 4
PF and global QoL scores during this period was assessed by a linear mixed model with fixed 5
effects for time as second-order polynomial to capture possible non-linear behavior. Random 6
effects for patients nested within cancer clinics were included to account for within-patient 7
correlations due to repeated measurements and possible within-clinic cluster effect.
8 9
To investigate if the frailty indicators and symptoms were associated to the patients’ overall 10
level of PF and global QoL during six months of follow-up, the linear mixed models were 11
adjusted for the frailty indicators and symptoms by first including them one by one into 12
bivariate models. Next, three multiple linear mixed models (A, B and C) for each outcome 13
were estimated. The independent impact of the frailty indicators was assessed by first 14
including them all into a multiple model (A). Then, model A was adjusted for age, gender, 15
and cancer related factors i.e. PS, type of cancer, stage of disease and treatment (model B).
16
Finally, the impact of symptom occurrence was investigated by adding symptom scores 17
reported simultaneously with PF and global QoL from baseline to six months to the model 18
(C). In each multiple model (A, B, C), all covariates were included simultaneously. As basic 19
ADL was derived from one item of the QLQ-C30 PF scale, which was also the outcome, this 20
frailty indicator was excluded from all models for PF. No co-linearity issues were detected 21
when performing correlation analysis.
22 23
The linear mixed model described above assesses the overall course of PF and global QoL 24
during six months for all patients. By means of an exploratory approach, growth mixture 25
9
model was estimated to identify possible unobserved groups of patients following distinct 1
trajectories in the main endpoint, PF. The method assesses individual trajectories and attempts 2
to group the patients with similar profiles together. The optimal number of groups was 3
determined by using Akaike’s Information Criterion (AIC) and aiming at average within- 4
group probabilities larger than 0.8, non-overlapping 95% CI for each trajectory, and 5
reasonable group size. The model does not include patient characteristics, thus identified 6
groups were next described by bivariate and multiple nominal regression models with group 7
membership as dependent variable and baseline characteristics as covariates. The included 8
covariates were age, gender, cancer related factors as described above, and baseline symptom 9
scores (pain, dyspnea, appetite loss, sleeping disturbances, constipation and nausea/vomiting).
10
AIC was used to reduce the multiple model for excessive variables.
11 12
The analyses were performed using SPSS v25 and STATA v14. Results with p-values below 13
0.05 were considered statistically significant.
14 15
Ethics 16
The study was approved by the Regional Committee for Medical and Health Research Ethics 17
South East Norway and registered at ClinicalTrials.gov (NCT01742442). All patients 18
provided written informed consent.
19 20
Results
21
Between January 2013 and April 2015, 307 patients were included (24). One patient withdrew 22
consent and 18 had missing baseline questionnaires. Thus, 288 (94%) patients were eligible 23
for this study. Mean age was 76.9 years, 56% were male, the majority had distant metastases 24
10
(56%) and received palliative treatment (68%) (Table 2). The patients reported a mean of 2.7 1
comorbidities and 4.1 daily medications, 15% were diagnosed as severely malnourished, 3%
2
had experienced more than one fall during the last six months, and the median (min-max) 3
GDS and MMSE scores were 2.0 (0-13) and 29 (19-30), respectively. At two, four, and six 4
months of follow-up, 13 (5%), 27 (9%), and 52 (18%) patients had died. The proportion of 5
completed QLQ-C30 questionnaires ranged from 89% to 95% of those alive at these time 6
points. Mean baseline PF, global QoL and symptom scores are shown in Table 2.
7 8
Impact of frailty indicators and symptoms on the overall level of PF and global QoL 9
According to unadjusted linear mixed models, assessing the overall course during follow-up, 10
PF declined non-linearly and statistically significantly (max 8.9 points at four months, 11
p<0.001), whereas the global QoL declined linearly (max 3.9 points at six months, p=0.008).
12
However, neither decline was clinically significant (Figure 1A and 1B).
13 14
Bivariate linear mixed models showed that all frailty indicators were significantly associated 15
with the patients’ overall level of PF during follow-up, as were also age, PS, type of cancer, 16
stage of disease, treatment, and all symptom scores measured simultaneously with PF (Table 17
3). In the multiple model including all frailty indicators, higher GDS-scores, poorer TUG, and 18
malnutrition were significantly associated with a poorer PF level within the study period 19
(Table 3, model A). In addition to PS, these factors were also the only significant covariates 20
when controlling for age, gender and the cancer-related factors (Table 3, model B). In the 21
final model (C), GDS-scores and TUG remained independent, significant covariates. Higher 22
scores on pain, dyspnea, appetite loss, and sleep disturbance throughout follow-up were also 23
significantly and independently associated with a poorer overall level of PF (Table 3, model 24
25 C).
11 1
Results of the corresponding analyses for global QoL are displayed in Table 4. In bivariate 2
linear mixed models, all frailty indicators except for basic ADL, number of falls, and MMSE, 3
were associated with the patients’ global QoL level during follow-up (p<0.01). According to 4
the multiple model A, malnutrition (p=0.004), higher GDS-score (p<0.001), poorer TUG 5
(p=0.013), and no ADL deficits (p=0.048) were independently associated with a poorer global 6
QoL level. When controlling for age, gender and cancer-related factors (model B), 7
malnutrition (p=0.013), GDS score (p<0.001), and TUG (p=0.041) remained the only 8
significant covariates. In model C, including patients’ symptom reports during follow-up, 9
higher GDS-scores (p<0.001), poorer TUG (p=0.029), more pain, dyspnea, appetite loss, 10
sleeping disturbances (all p<0.001), and diarrhea (p=0.018) were significantly associated with 11
poorer overall global QoL level throughout the study period (Table 4).
12 13
Trajectory analyses to identify distinct subgroups of PF development 14
Growth mixture model identified three groups of patients with distinct PF trajectories i.e. poor 15
(n=69, 24%), intermediate (n=103, 36%), and good (n=112, 40%) with high mean within- 16
group probabilities (Table 5) and non-overlapping 95% CI (Figure 1C). The poor group had a 17
significantly poorer mean PF score at baseline (mean 51.6 SD 20.8) compared to the 18
intermediate (68.3, SD 13.7) and good (91.5, SD 9.5) groups, and exhibited a non-linear 19
statistically and clinically significant decline by 20.2 points over four months (p<0.001). The 20
good group remained stable throughout the follow-up period, and the intermediate group 21
experienced a statistically, though not clinically significant linear decrease (p=0.003) (Table 22
5, Figure 1C).
23 24
12
For all frailty indicators and baseline symptom scores, more deficits and higher symptom 1
intensity were registered for the poor PF group in comparison to the intermediate group, 2
which in turn had more deficits and reported more symptoms than the good PF group (Table 3
2). According to bivariate nominal regression models, the poor and good PF groups differed 4
significantly on all the considered covariates except for number of falls, gender, and diarrhea 5
(data not shown). In the AIC-reduced multiple model, higher GDS-scores, poorer TUG, and 6
more pain and dyspnea were significantly and independently associated with higher odds of 7
belonging to the poor PF group as compared to good group (OR 1.3 (1.1; 1.5), p=0.008; OR 8
1.8 (1.5; 2.2), p<0.001; OR 1.0 (1.0; 1.1); p<0.001, and OR 1.0 (1.0; 11.1), p<0.001, 9
respectively).
10 11
Within six months, there were also differences in survival between the groups. Whereas 62%
12
of the patients in the poor group survived for six months, the corresponding percentages in the 13
intermediate and good groups were 84% and 92%, respectively (p<0.001).
14 15
Discussion
16
In the present study of older patients referred for systemic cancer treatment, we showed that 17
pre-treatment higher GDS and poorer TUG scores were independently associated to poorer 18
overall levels of patient-reported PF and global QoL during six months of follow-up.
19
Furthermore, more pain, dyspnea, appetite loss, and sleep disturbances within the same period 20
had a profoundly negative impact on both outcomes. Pre-treatment malnutrition was also 21
associated with poorer PF and global QoL scores, although not independently of symptom 22
scores. Exploratory analyses identified three groups of patients with distinct PF trajectories.
23
The poor PF group, comprising 24% of the patients, had the poorest PF at baseline and 24
13
reported a clinically significant decline during the study period. In line with our main 1
findings, belonging to this group was independently associated with higher GDS and poorer 2
TUG scores, more pain, and dyspnea at baseline.
3 4
We are not aware of any former studies reporting how individual frailty indicators may be 5
associated with global QoL in older patients during systemic cancer treatment, or 6
investigating the longitudinal relationship between symptoms, physical function, and QoL in 7
such patient cohorts. The negative effect of symptom distress found in our study is, however, 8
in line with several cross-sectional studies describing correlations between symptom severity, 9
impairments in physical function, and QoL (21-23). Three recent studies have investigated if 10
pre-treatment GA elements may be associated with functional decline in terms of reduced 11
ability to carry out daily life activities. Decoster et al. reported no independent impact of any 12
of these frailty indicators in newly diagnosed patients with lung cancer (4). Hoppe et al (5) 13
and Kenis et al (3), both studying patients with various cancer types receiving chemotherapy, 14
found that impairments in instrumental ADL (IADL), higher GDS-scores and malnutrition 15
predicted declining ADL. Their results may not be directly comparable to ours due to 16
differences in assessment tool and methods. Whereas we used patient-report, their 17
assessments were made by a geriatrician or a trained nurse, and these measures may only be 18
moderately correlated. Jointly, however, the studies strongly indicate a substantial negative 19
impact of pre-treatment physical impairments, depressive symptoms, and malnutrition on 20
older patients' physical function during cancer treatment. According to our findings, the same 21
factors are of major importance for global QoL.
22 23
The proportion of patients experiencing a decline in physical function in our study was 24
consistent with several other reports on older patients with cancer (3-6). A recent study also 25
14
identified three patient groups with distinct trajectories of patient-reported physical function, 1
i.e. poor, intermediate, and good (38), though these were all stable. Supporting our finding, 2
depression, and lower physical activity were among the main characteristics within the poor 3
group. Moreover, it is worth noting that PF scores in our good PF group were higher than 4
reported in a Norwegian reference population, 70 - 79 years of age (female scores 74.9, male 5
scores 84.2) (39). Baseline scores for the poor PF group were comparable to those found in a 6
cancer population with expected survival of three months (scores 46-48) (40), indicating that 7
the observed decline of 20 points may have serious implications for the patients.
8 9
The dismal consequences of physical impairment, depression, and malnutrition for cancer 10
survival and treatment complications are well known (41-46).Our findings extend this 11
knowledge, indicating that such problems should also be properly addressed in order to 12
maintain older patients’ physical function and QoL throughout systemic cancer treatment.
13
Pre-habilitation and rehabilitation programs including physical exercise and/or nutritional 14
interventions have proven successful in other settings, also among palliative patients (47, 48).
15
Exploring the reasons for depression might be equally important. Motivational and neuro- 16
hormonal mechanisms may for example underlie the association between depression and 17
decline in physical function, and pharmacotherapy and cognitive-behavioral interventions 18
might be helpful (49).
19 20
The significant, negative associations between symptom distress during the disease course 21
and patients’ PF and global QoL scores reinforce the need to follow patients with systematic 22
and repeated symptom assessment. Despite being highly recommended, this is seldom 23
routinely applied, and is cited as a major reason for inadequate symptom management (50).
24
Consistent with this, evidence is emerging suggesting that systematic symptom monitoring 25
15
using patient-reported outcome measures followed by targeted interventions may improve 1
cancer patients’ outcomes, including QoL and survival (51, 52). The present study provides 2
no information on treatment response and one might therefore argue that the associations 3
between poorer PF and global QoL scores and more symptoms may reflect cancer 4
progression. It should, however, be noted that even in the group with the poorest trajectory of 5
PF scores, the majority lived for more than six months. Thus, early decline in physical 6
function, poor QoL and a high symptom burden should not be seen as inevitable, but acted 7
upon. For older patients, however, physical symptoms as well as physical impairment, 8
depression, and malnutrition are most likely multifactorial due to co-existing problems.
9
Hence, interventions aiming at maintaining physical function and QoL should be 10
individualized and based on GA in accordance with current recommendations (53).
11 12
Our study has several limitations. Firstly, we included a heterogeneous sample of patients 13
with several different cancer diagnoses, stages and treatment. Secondly, the choice of 14
assessment tools may have impacted our results. This particularly applies to our comorbidity 15
assessment, since comorbidity has been found to affect older patients’ physical function and 16
QoL in other studies using more comprehensive assessments than the OARS (38, 54).Thirdly, 17
the multitude of factors included in our analyses may introduce uncertainties, and the 18
exploratory analysis related to PF trajectories should be interpreted with caution. Fourthly, it 19
may be argued that fatigue, which is a symptom that may seriously affect patients’ physical 20
function and QoL, should have been taken into account. However, fatigue has no uniform, 21
established treatment, and most treatment strategies include treatment of possibly contributing 22
factors, such as malnutrition, depression, pain, and sleep disturbances (55, 56). Consequently, 23
we defined that including the fatigue scores in our analyses would be of little benefit since our 24
analyses comprised a wide range of factors that may contribute to fatigue and be efficiently 25
16
treated if properly assessed and detected. Thus, systematically targeting the problems found to 1
affect PF and global QoL in our study may also improve fatigue (56), which would be an 2
important additional outcome in studies aiming to evaluate such an approach.
3 4
Strengths of our study are the relatively large sample size, and that factors taken into account 5
were predefined based on former studies and clinical judgement. Our frailty indicators 6
covered recommended domains (16, 18), and were assessed by validated instruments. The 7
QLQ-C30, used for outcome and symptom assessment, provides high completion rates, is 8
widely applied and validated, sensitive to change, and is a recommended measure of physical 9
function (57). Compared to performance measures, patient-reported physical function has 10
been found to have similar psychometric properties, and as patient-report reflects patients’
11
experience from routine life, such measures may also more appropriately capture factors that 12
affect their day to day function (58). In a longitudinal study, however, one can never rule out 13
that a potential response shift, i.e. a psychological adaptation to changing health status, may 14
have occurred. From an observational point of view, declines in physical function and QoL 15
may therefore have been more profound than what was reflected by the patients’ scores.
16 17
In conclusion, pre-treatment physical impairments, nutritional deficits, depressive and somatic 18
symptoms are associated with poor physical function and global QoL during the course of 19
disease in older patients with cancer, as is also unrelieved symptom distress within the same 20
period. Systematic symptom assessments and interventions targeted to these specific areas 21
might improve these outcomes. Further research is urgently needed to evaluate the effect and 22
feasibility of such interventions, and to provide more information on the course of physical 23
function and QoL during cancer therapy that may be used to facilitate treatment decisions.
24
17
Preferably, these studies should include homogeneous cohorts in terms of diagnosis, stage, 1
and treatment, and appropriately assess treatment response and side effects.
2 3
Individual authors’ contribution 4
Study concepts: MS Jordhøy, G Selbæk.
5 6
Study design: MS Jordhøy, G Selbæk, TB Wyller, MJ Hjermstad, S Rostoft 7
8
Data acquisition: L Kirkhus, M Harneshaug, MS Jordhøy.
9 10
Quality control of data and algorithms: L Kirkhus, M Harneshaug, J Šaltytė Benth, MS 11
Jordhøy.
12 13
Data analysis and interpretation: L Kirkhus, M Harneshaug, J Šaltytė Benth, MS Jordhøy, 14
S Rostoft, BH Grønberg 15
16
Statistical analysis: J Šaltytė Benth.
17 18
Manuscript preparation: L Kirkhus, M Harneshaug, J Šaltytė Benth, MS Jordhøy, S 19
Rostoft, BH Grønberg, G Selbæk, TB Wyller, MJ Hjermstad, S Bergh, Ø Kirkevold, T Borza, 20
I Saltvedt 21
22
Manuscript editing: L Kirkhus, M Harneshaug, J Šaltytė Benth, MS Jordhøy, S Rostoft, BH 23
Grønberg, G Selbæk, TB Wyller, MJ Hjermstad, S Bergh, Ø Kirkevold, T Borza, I Saltvedt 24
25
Manuscript review: L Kirkhus, M Harneshaug, J Šaltytė Benth, MS Jordhøy, S Rostoft, BH 26
Grønberg, G Selbæk, TB Wyller, MJ Hjermstad, S Bergh, Ø Kirkevold, T Borza, I Saltvedt 27
28 29
Acknowledgement 30
31
This study was funded by Innlandet Hospital Trust and registered at ClinicalTrials.gov 32
(NCT01742442). We want to thank the cancer clinics at Innlandet Hospital Trust, Oslo 33
University Hospital (OUH) and Akershus University Hospital (AHUS) for their participation 34
in the study. A special thanks to the study nurses at all locations who participated in the 35
inclusion and assessment of patients, and to the local principal investigators at OUH and 36
AHUS: Morten Brændengen and Olav Yri.
37 38
Declaration of conflicting interests 39
Dr. Gronberg reports grants from MSD, Roche, AstraZeneca, BMS, Pfizer, personal fees 40
from Takeda, MSD, Roche, AstraZeneca, BMS, Pfizer, Eli Lilly, Bayer, Pierre Fabre, 41
Novartis, Boehringer Ingelheim, grants from Roche, outside the submitted work; . Dr.
42
Saltvedt reports research collaboration with Boehringer Ingelheim, outside the submitted 43
18
work. The rest of the authors declare no potential conflicts of interest with respect to the 1
research, authorship or publications of this article.
2 3
References 4
1. Covinsky KE, Palmer RM, Fortinsky RH, Counsell SR, Stewart AL, Kresevic D, et al. Loss of 5
independence in activities of daily living in older adults hospitalized with medical illnesses: increased 6
vulnerability with age. J Am Geriatr Soc. 2003;51(4):451-8.
7
2. Derks MG, de Glas NA, Bastiaannet E, de Craen AJ, Portielje JE, van de Velde CJ, et al. Physical 8
Functioning in Older Patients With Breast Cancer: A Prospective Cohort Study in the TEAM Trial.
9
Oncologist. 2016;21(8):946-53.
10
3. Kenis C, Decoster L, Bastin J, Bode H, Van Puyvelde K, De Greve J, et al. Functional decline in 11
older patients with cancer receiving chemotherapy: A multicenter prospective study. J Geriatr Oncol.
12
2017;8(3):196-205.
13
4. Decoster L, Kenis C, Schallier D, Vansteenkiste J, Nackaerts K, Vanacker L, et al. Geriatric 14
Assessment and Functional Decline in Older Patients with Lung Cancer. Lung. 2017;195(5):619-26.
15
5. Hoppe S, Rainfray M, Fonck M, Hoppenreys L, Blanc JF, Ceccaldi J, et al. Functional decline in 16
older patients with cancer receiving first-line chemotherapy. J Clin Oncol. 2013;31(31):3877-82.
17
6. van Abbema D, van Vuuren A, van den Berkmortel F, van den Akker M, Deckx L, Buntinx F, et 18
al. Functional status decline in older patients with breast and colorectal cancer after cancer 19
treatment: A prospective cohort study. J Geriatr Oncol. 2017;8(3):176-84.
20
7. Mogal HD, Howard-McNatt M, Dodson R, Fino NF, Clark CJ. Quality of life of older African 21
American breast cancer survivors: a population-based study. Support Care Cancer. 2017;25(5):1431- 22
8.
23
8. Deschler B, Ihorst G, Hull M, Baier P. Regeneration of older patients after oncologic surgery.
24
A temporal trajectory of geriatric assessment and quality of life parameters. J Geriatr Oncol.
25
2019;10(1):112-9.
26
9. Fried TR, Bradley EH, Towle VR, Allore H. Understanding the treatment preferences of 27
seriously ill patients. N Engl J Med. 2002;346(14):1061-6.
28
10. Salkeld G, Cameron ID, Cumming RG, Easter S, Seymour J, Kurrle SE, et al. Quality of life 29
related to fear of falling and hip fracture in older women: a time trade off study. BMJ.
30
2000;320(7231):341-6.
31
11. Yellen SB, Cella DF, Leslie WT. Age and clinical decision making in oncology patients. J Natl 32
Cancer Inst. 1994;86(23):1766-70.
33
12. Celis ESPD, Li D, Sun C-L, Kim H, Twardowski P, Fakih M, et al. Patient-defined goals and 34
preferences among older adults with cancer starting chemotherapy (CT). Journal of Clinical Oncology.
35
2018;36(15_suppl):10009-.
36
13. Bluethmann SM, Mariotto AB, Rowland JH. Anticipating the "Silver Tsunami": Prevalence 37
Trajectories and Comorbidity Burden among Older Cancer Survivors in the United States. Cancer 38
Epidemiol Biomarkers Prev. 2016;25(7):1029-36.
39
14. Clegg A, Young J, Iliffe S, Rikkert MO, Rockwood K. Frailty in elderly people. Lancet.
40
2013;381(9868):752-62.
41
15. Henchoz Y, Bula C, Guessous I, Santos-Eggimann B. Association between Physical Frailty and 42
Quality of Life in a Representative Sample of Community-Dwelling Swiss Older People. J Nutr Health 43
Aging. 2017;21(5):585-92.
44
16. Wildiers H, Heeren P, Puts M, Topinkova E, Janssen-Heijnen ML, Extermann M, et al.
45
International Society of Geriatric Oncology consensus on geriatric assessment in older patients with 46
cancer. J Clin Oncol. 2014;32(24):2595-603.
47
19
17. Magnuson A, Allore H, Cohen HJ, Mohile SG, Williams GR, Chapman A, et al. Geriatric 1
assessment with management in cancer care: Current evidence and potential mechanisms for future 2
research. J Geriatr Oncol. 2016.
3
18. Mohile SG, Dale W, Somerfield MR, Schonberg MA, Boyd CM, Burhenn PS, et al. Practical 4
Assessment and Management of Vulnerabilities in Older Patients Receiving Chemotherapy: ASCO 5
Guideline for Geriatric Oncology. J Clin Oncol. 2018;36(22):2326-47.
6
19. Puts MT, Santos B, Hardt J, Monette J, Girre V, Atenafu EG, et al. An update on a systematic 7
review of the use of geriatric assessment for older adults in oncology. Ann Oncol. 2014;25(2):307-15.
8
20. Ronning B, Wyller TB, Jordhoy MS, Nesbakken A, Bakka A, Seljeflot I, et al. Frailty indicators 9
and functional status in older patients after colorectal cancer surgery. J Geriatr Oncol. 2014;5(1):26- 10
32.
11
21. Pandya C, Magnuson A, Flannery M, Zittel J, Duberstein P, Loh KP, et al. Association Between 12
Symptom Burden and Physical Function in Older Patients with Cancer. J Am Geriatr Soc.
13
2019;67(5):998-1004.
14
22. Kurtz ME, Kurtz JC, Stommel M, Given CW, Given BA. Symptomatology and loss of physical 15
functioning among geriatric patients with lung cancer. J Pain Symptom Manage. 2000;19(4):249-56.
16
23. Cheng KK, Yeung RM. Symptom distress in older adults during cancer therapy: impact on 17
performance status and quality of life. J Geriatr Oncol. 2013;4(1):71-7.
18
24. Kirkhus L, Saltyte Benth J, Rostoft S, Gronberg BH, Hjermstad MJ, Selbaek G, et al. Geriatric 19
assessment is superior to oncologists' clinical judgement in identifying frailty. Br J Cancer.
20
2017;117(4):470-7.
21
25. Kirkhus L, Saltyte Benth J, Gronberg BH, Hjermstad MJ, Rostoft S, Harneshaug M, et al. Frailty 22
identified by geriatric assessment is associated with poor functioning, high symptom burden and 23
increased risk of physical decline in older cancer patients: Prospective observational study. Palliat 24
Med. 2019:269216319825972.
25
26. Aaronson NK, Ahmedzai S, Bergman B, Bullinger M, Cull A, Duez NJ, et al. The European 26
Organization for Research and Treatment of Cancer QLQ-C30: a quality-of-life instrument for use in 27
international clinical trials in oncology. J Natl Cancer Inst. 1993;85(5):365-76.
28
27. Fayers P, Aaronson NK, Bjordal K, Groenvold M, D C, A B, et al. The EORTC QLQ-C30 Scoring 29
Manual (3rd Edition). European Organisation for Research and Treatment of Cancer, Brussels; 2001.
30
28. Osoba D, Rodrigues G, Myles J, Zee B, Pater J. Interpreting the significance of changes in 31
health-related quality-of-life scores. J Clin Oncol. 1998;16(1):139-44.
32
29. Snyder CF, Blackford AL, Sussman J, Bainbridge D, Howell D, Seow HY, et al. Identifying 33
changes in scores on the EORTC-QLQ-C30 representing a change in patients' supportive care needs.
34
Qual Life Res. 2015;24(5):1207-16.
35
30. Balducci L, Extermann M. Management of cancer in the older person: a practical approach.
36
Oncologist. 2000;5(3):224-37.
37
31. Hurria A, Gupta S, Zauderer M, Zuckerman EL, Cohen HJ, Muss H, et al. Developing a cancer- 38
specific geriatric assessment: a feasibility study. Cancer. 2005;104(9):1998-2005.
39
32. Fillenbaum GG, Smyer MA. The development, validity, and reliability of the OARS 40
multidimensional functional assessment questionnaire. J Gerontol. 1981;36(4):428-34.
41
33. Ottery FD. Definition of standardized nutritional assessment and interventional pathways in 42
oncology. Nutrition. 1996;12(1 Suppl):S15-9.
43
34. Yesavage JA BT, Rose TL, Lum O, Huang V, Adey M, et al. Development and validation of a 44
geriatric depression screening scale: a preliminary report. J Psychiatr Res. 1982;17(1):12.
45
35. Folstein MF, Folstein SE, McHugh PR. "Mini-mental state". A practical method for grading the 46
cognitive state of patients for the clinician. J Psychiatr Res. 1975;12(3):189-98.
47
36. Podsiadlo D, Richardson S. The timed "Up & Go": a test of basic functional mobility for frail 48
elderly persons. J Am Geriatr Soc. 1991;39(2):142-8.
49
37. Beauchet O, Fantino B, Allali G, Muir SW, Montero-Odasso M, Annweiler C. Timed Up and Go 50
test and risk of falls in older adults: a systematic review. J Nutr Health Aging. 2011;15(10):933-8.
51
20
38. Miaskowski C, Wong ML, Cooper BA, Mastick J, Paul SM, Possin K, et al. Distinct Physical 1
Function Profiles in Older Adults Receiving Cancer Chemotherapy. J Pain Symptom Manage.
2
2017;54(3):263-72.
3
39. Fossa SD, Hess SL, Dahl AA, Hjermstad MJ, Veenstra M. Stability of health-related quality of 4
life in the Norwegian general population and impact of chronic morbidity in individuals with and 5
without a cancer diagnosis. Acta Oncol. 2007;46(4):452-61.
6
40. Jordhoy MS, Fayers P, Loge JH, Ahlner-Elmqvist M, Kaasa S. Quality of life in palliative cancer 7
care: results from a cluster randomized trial. J Clin Oncol. 2001;19(18):3884-94.
8
41. Maltoni M, Caraceni A, Brunelli C, Broeckaert B, Christakis N, Eychmueller S, et al. Prognostic 9
factors in advanced cancer patients: evidence-based clinical recommendations--a study by the 10
Steering Committee of the European Association for Palliative Care. J Clin Oncol. 2005;23(25):6240-8.
11
42. Extermann M, Boler I, Reich RR, Lyman GH, Brown RH, DeFelice J, et al. Predicting the risk of 12
chemotherapy toxicity in older patients: the Chemotherapy Risk Assessment Scale for High-Age 13
Patients (CRASH) score. Cancer. 2012;118(13):3377-86.
14
43. Verweij NM, Schiphorst AH, Pronk A, van den Bos F, Hamaker ME. Physical performance 15
measures for predicting outcome in cancer patients: a systematic review. Acta Oncol. 2016:1-6.
16
44. Arends J, Baracos V, Bertz H, Bozzetti F, Calder PC, Deutz NEP, et al. ESPEN expert group 17
recommendations for action against cancer-related malnutrition. Clin Nutr. 2017;36(5):1187-96.
18
45. Barber B, Dergousoff J, Slater L, Harris J, O'Connell D, El-Hakim H, et al. Depression and 19
Survival in Patients With Head and Neck Cancer: A Systematic Review. JAMA otolaryngology-- head &
20
neck surgery. 2016;142(3):284-8.
21
46. Sullivan DR, Forsberg CW, Ganzini L, Au DH, Gould MK, Provenzale D, et al. Longitudinal 22
Changes in Depression Symptoms and Survival Among Patients With Lung Cancer: A National Cohort 23
Assessment. J Clin Oncol. 2016;34(33):3984-91.
24
47. Gillis C, Buhler K, Bresee L, Carli F, Gramlich L, Culos-Reed N, et al. Effects of Nutritional 25
Prehabilitation, With and Without Exercise, on Outcomes of Patients Who Undergo Colorectal 26
Surgery: A Systematic Review and Meta-analysis. Gastroenterology. 2018;155(2):391-410.e4.
27
48. Salakari MR, Surakka T, Nurminen R, Pylkkanen L. Effects of rehabilitation among patients 28
with advances cancer: a systematic review. Acta Oncol. 2015;54(5):618-28.
29
49. Parpa E, Tsilika E, Gennimata V, Mystakidou K. Elderly cancer patients' psychopathology: a 30
systematic review: aging and mental health. Arch Gerontol Geriatr. 2015;60(1):9-15.
31
50. Kaasa S, Loge JH, Aapro M, Albreht T, Anderson R, Bruera E, et al. Integration of oncology and 32
palliative care: a Lancet Oncology Commission. Lancet Oncol. 2018;19(11):e588-e653.
33
51. Basch E, Deal AM, Kris MG, Scher HI, Hudis CA, Sabbatini P, et al. Symptom Monitoring With 34
Patient-Reported Outcomes During Routine Cancer Treatment: A Randomized Controlled Trial. J Clin 35
Oncol. 2016;34(6):557-65.
36
52. Basch E, Deal AM, Dueck AC, Scher HI, Kris MG, Hudis C, et al. Overall Survival Results of a 37
Trial Assessing Patient-Reported Outcomes for Symptom Monitoring During Routine Cancer 38
Treatment. JAMA. 2017;318(2):197-8.
39
53. Comprehensive geriatric assessment for patients with cancer [Internet]. UpToDate. 2019 40
[cited 16.02.19].
41
54. Wedding U, Rohrig B, Klippstein A, Brix C, Pientka L, Hoffken K. Co-morbidity and functional 42
deficits independently contribute to quality of life before chemotherapy in elderly cancer patients.
43
Support Care Cancer. 2007;15(9):1097-104.
44
55. Mohandas H, Jaganathan SK, Mani MP, Ayyar M, Rohini Thevi GV. Cancer-related fatigue 45
treatment: An overview. J Cancer Res Ther. 2017;13(6):916-29.
46
56. de Raaf PJ, de Klerk C, Timman R, Busschbach JJ, Oldenmenger WH, van der Rijt CC.
47
Systematic monitoring and treatment of physical symptoms to alleviate fatigue in patients with 48
advanced cancer: a randomized controlled trial. J Clin Oncol. 2013;31(6):716-23.
49
57. Atkinson TM, Stover AM, Storfer DF, Saracino RM, D'Agostino TA, Pergolizzi D, et al. Patient- 50
Reported Physical Function Measures in Cancer Clinical Trials. Epidemiol Rev. 2017;39(1):59-70.
51
21
58. Latham NK, Mehta V, Nguyen AM, Jette AM, Olarsch S, Papanicolaou D, et al. Performance- 1
based or self-report measures of physical function: which should be used in clinical trials of hip 2
fracture patients? Arch Phys Med Rehabil. 2008;89(11):2146-55.
3 4
Table 1 Overview of frailty indicators (as a part of the modified geriatric assessment) performed at patient inclusion
Domain Assessment Rated by Variable name Scores and ranges Interpretation
Comorbidity The Physical Health Section of the Older Americans’ Resources and Services comorbidity scale (OARS)
Patient Numberof comorbidities 0-15 (continuous)
Medication Nurse Numberof medications (continuous)
Nutritional status Patient-generated Subjective Global Assessment (PG-SGA)
Nurse/patient Malnutrition Yes=Considered severely malnourished by nurse or self-reported weight loss of
≥10% the last 6 months No=None of the above
Depressive symptoms 15-item Geriatric depression scale (GDS-15) Patient GDS 0-15 (continuous) Higher scores =
more symptoms Cognitive function Norwegian Revised Mini Mental State
Examination (NR-MMSE)
Nurse MMSE 0-30 (continuous) Higher scores =
better function Falls the last six
months
Nurse Numberof falls 0-1 or ≥ 2
Mobility Timed Up and Go test (TUG) (fast pace) Nurse TUG number of seconds (continuous)
Activities of daily living (ADL)
Question no. 5 from the physical functioning scale on the European Organisation for Research and Treatment of Cancer Quality of Life Core Questionnaire-C30
Patient ADL: “Do you need help with eating, dressing, washing yourself or using the toilet?” (dichotomized)
Yes ="A little", "some" or "very much"
or No ="Not at all"
Table 2. Baseline characteristics of the entire cohort (N = 288) and of three patient groups with distinct trajectories of physical function (N=284)
Characteristics
Physical function trajectory All patients
(n=288)
Poor (n=69)
Intermediate (n=103)
Good (n=112) Age, mean (SD) 76.9 (5.1) 78.1 (5.5) 76.9 (5.2) 76.0 (4.7)
Female gender, n (%) 126 (44) 33 (48) 46 (45) 44 (39)
Cancer type, n (%)
Colorectal 83 (29) 15 (22) 27 (26) 38 (34)
Lung 59 (21) 25 (36) 21 (20) 13 (12)
Prostate 56 (19) 10 (14) 24 (23) 22 (20)
Other gastrointestinal 34 (12) 6 (9) 14 (14) 14 (12)
Breast 30 (10) 4 (6) 6 (6) 19 (17)
Other 26 (9) 9 (13) 11 (11) 6 (5)
Stage, n (%)
Localized 73 (25) 11 (16) 25 (24) 35 (32)
Locally advanced 55 (19) 10 (14) 21 (20) 24 (21)
Metastatic 160 (56) 48 (70) 57 (56) 53 (47)
Treatment, n (%)
Curative 91 (32) 10 (14) 31 (30) 48 (43)
Palliative chemotherapy 126 (44) 40 (58) 45 (44) 39 (35)
Other palliative systemic
cancer treatment 51 (18) 8 (12) 24 (23) 19 (17)
Other palliative care 20 (7) 11 (16) 3 (3) 6 (5)
ECOG PSa 2-4, n (%) 43 (15) 25 (36) 13 (13) 5 (5)
Number of comorbidities,
mean (SD) 2.7 (1.7) 3.2 (2.0) 3.1 (1.7) 2.2 (1.4)
Number of medications,
mean (SD) 4.1 (2.9) 4.9 (3.2) 4.8 (2.9) 3.1 (2.4)
Malnutrition, n (%) 43 (15) 19 (28) 16 (16) 7 (6)
GDSb score, mean (SD) 2.9 (2.8) 4.5 (3.1) 3.3 (2.8) 1.6 (2.0)
> 2 falls last six months, n
(%) 10 (3) 5 (7) 4 (4) 1 (1)
MMSEc score, mean (SD) 28.5 (1.9) 27.9 (2.1) 28.5 (2.1) 28.9 (1.5) TUGd seconds, mean (SD) 8.7 (3.5) 11.2 (4.5) 9.3 (3.3) 6.9 (1.7) EORTC QLQ C30e scores,
mean (SD)
Physical function 72.9 (21.4) 51.6 (20.8) 68.3 (13.7) 91.5 (9.5)
Global QoL 64.1 (23.1) 51.0 (22.6) 56.9 (19.8) 79.0 (17.3)
Pain 24.8 (29.4) 42.5 (34.1) 301 (28.2) 9.4 (17.6)
Dyspnoea 25.7 (31.4) 41.1 (36.7) 29.1 (32.7) 13.4 (20.2)
Appetite loss 21.4 (31.4) 35.7 (37.2) 24.9 (32.2) 9.8 (21.3)
Constipation 24.0 (29.3) 36.7 (35.8) 28.5 (28.9) 12.5 (20.1)
Sleeping disturbance 26.2 (28.5) 38.2 (31.5) 26.8 (27.0) 18.2 (25.3)
Diarrhea 15.2 (22.4) 16.4 (23.3) 14.6 (22.2) 14.5 (21.9)
a Eastern Cooperative Oncology Group Performance status b 15-item Geriatric depression scale
c Norwegian Revised Mini Mental State Examination dTimed Up and Go test eEuropean Organization for Research and Treatment of Cancer Core Quality of_Life Questionnaire
